Project Summary Alcohol tax increases are known to increase prices and reduce alcohol consumption. However, little is known about the specifics of this process. This project will investigate the details of how an excise tax increase affects the alcohol market. The overarching goal of this research is to determine if heavy drinkers use substitution options to avoid the effects of a tax increase to a greater degree than moderate and low- income drinkers. The market changes on the supply side because not all products are required to increase prices by the same amount or proportion. On the demand side, buyers may engage in a series of avoidance strategies. These include buying lower quality and thus cheaper products, buying from different discount retailers, and buying larger bottle sizes. Buying less total alcohol is another option. Heavy drinkers have an incentive to incur greater search and travel cost than moderate drinkers because these costs are independent of the amount purchased, while benefits are positively related to consumption. Two large exogenous state tax changes will be studied. These changes were in Illinois and in Washington State. These changes are considered quasi-natural experiments because they rely on assignment to treatment using a criterion other than random assignment. The analytic approach will be a series of difference-in-differences models. The Nielsen Homescan data will be used to investigate the characteristics of buyers of low priced alcohol. It is of particular interest to investigate how much low priced alcohol is bought by heavy drinkers relative to low- income moderate drinkers. The Nielsen Retail Scanner data set provides detailed store prices. The retail price data will be used to study the change in prices due to the tax increase at different points in the price distribution. This is important to verify that relative prices change when excise taxes increase. Detailed data on purchases by store type, bottle size and brand are also available in the Nielsen Homescan data set. The Homescan data will be used to investigate substitutions to avoid the price increase that occurs after the tax increase by heavy versus moderate drinkers and by low-income drinkers relative to higher-income drinkers. The Homescan and Behavioral Risk Factor Surveillance System data will also be used to investigate changes in consumption by heavy drinkers relative to moderate drinkers as a result of the excise tax increases. This set of goals will create a detailed view of who buys low priced alcohol, the impact of the tax on relative alcohol prices and the price increase avoidance behavior of heavy drinkers when excise taxes increase. If this project shows that heavy drinkers rely on substitutions to minimize the impact of an excise tax increase, it would suggest that price increases at the low end of the distribution are more important than increases in the average price. This type of tax avoidance could be limited with minimum pricing policies. Alternatively, if this project finds that the impact of excise tax increases is not avoided by heavy drinkers then this information would be useful in future discussions of excise tax increases.